lines at the
back of the stone would alone serve to distinguish the two stones.

[Illustration: FIG. 2.]

A SIMPLE BUT VERY VALUABLE TEST FOR THE KIND OF REFRACTION OF A CUT
STONE. In the case of most of the other doubly refracting stones the
degree of separation is much less than in peridot and zircon, and it
takes a well-trained and careful eye to detect the doubling of the
lines. Here a very simple device will serve to assist the eye in
determining whether a cut stone is singly or doubly refracting. Expose
the stone to direct sunlight and hold an opaque white card a few inches
from the stone, in the direction of the sun, so as to get the bright
reflections _from within the stone_ reflected onto the card.

If the material is singly refractive (as in the case of diamond, garnet,
spinel, and glass), _single images_ of each of the reflecting facets
will appear on the card, but if doubly refracting--even if slightly
so--_double images_ will appear. When the stone is slightly moved,
these pairs of reflections will travel _together as pairs_ and not tend
to separate. The space between the two members of each pair of
reflections serves to give a rough idea of the degree of the double
refraction of the material if compared with the space between members in
the case of some other kind of stone held at the same distance from the
card. Thus zircon separates the reflections widely. Aquamarine, which is
feebly doubly refracting, separates them but slightly.

It will be seen at once that we have here a very easily applied test and
one that requires no costly apparatus. It is, furthermore, a sure test,
after a little practice. For example, if one has something that looks
like a fine emerald, but that may be glass, all one need to do is to
expose it in the sun, as above indicated. If real emerald, double images
will be had (very close together, because emerald is but feebly doubly
refracting). If glass, the images on the card will be single.

Similarly, ruby can at once be distinguished from even the finest garnet
or ruby spinel, as the last two are singly refracting. So, too, are
glass imitations of ruby and ruby doublets (which consist of glass and
garnet). This test cannot injure the stone, it may be applied to mounted
stones, and it is reliable. For stones of very deep color this test may
fail for lack of sufficiently brilliant reflections. In such a case hold
the card _beyond_ the stone and let the sunlight shine _through_